(P 361) MOLECULAR CHARACTERIZATION OF SOLITARY FIBROUS TUMORS BY ANATOMIC SITE AND NAB2::STAT6 FUSION VARIANT TYPE

Objective: Solitary Fibrous Tumor (SFT) is a rare mesenchymal tumor with an incidence of 1 new case/per million people/year. Advances in immunohistochemistry and molecular diagnostics identified STAT6 nuclear protein expression and the NAB2::STAT6 fusion gene as more precise tools for SFT diagnosis. Still, since the fusion transcript driver’s first description in 2013, few new insights have been brought into critical molecular events in SFT. The prognosis of SFT patients is associated with anatomic localization, and an independent prognostic impact of the NAB2::STAT6 fusion variant has been suggested. We aimed with our study to evaluate the molecular landscape of SFT patients and the impact of the NAB2::STAT6 fusion variant type on the largest cohort reported to date on this rare type of sarcoma.

Methods: This is an institutional review board approved, retrospective analysis of next-generation DNA (targeted, n=24, or whole exome, n=104) and RNA (whole transcriptome, n=128) sequencing data for patient tumors submitted to Caris Life Sciences (Phoenix, AZ). Included in this analysis were 128 samples with a reported diagnosis of SFT and NAB2::STAT6 fusion transcript detected, along with 16 NAB2::STAT6 fusion-positive samples with non-SFT histology reported. NAB2::STAT6 fusion variant were categorized based on associated fusion junction exons. Tissue biopsy sites were classified as thoracic (T), extra-thoracic (ET), or CNS according to anatomic location. Tumor microenvironment (TME) immune cell population abundance estimated by using MCP-counter (Becht, 2016). Statistical significance was determined using X2 and Mann-Whitney U tests.

Results: SFT samples were comprised of equal proportions of male and female patients (50%, n=64 each), with a median age of 61 years, and 59% (n=76) were metastatic biopsies (8 unknown). The most common NAB2::STAT6 fusion transcripts were Exon6::Exon16 (Ex6::Ex16; 41%, n=52), Ex4::Ex2 (27%, n=35), and Ex6::Ex17 (13%, n=16), with several other variant transcripts with low prevalence of fused exons (‘other’; 19%, n=25). Thoracic SFTs frequently harbored Ex4::Ex2 variant transcripts compared to non-thoracic tumors (78%, n=21; ET: 13%, n=7; CNS 14%, n=7; p< 0.001)t. SFTs from ET and CNS sites had similar enrichment of Ex6::Ex16 variants (ET: 48%, n=25; CNS: 47%, n=23; T: 15%, n=4; p< 0.001) and greater proportions of tumors with other rare variant transcripts compared to T sites (ET: 23%, n=12; CNS: 24%, n=12; T: 4%, n=1; p< 0.001). Aside from the NAB2::STAT6 fusion, tumor suppressor genes TP53 (28%, n=36), RB1 (5.4%, n=5), and PTEN (2.3%, n=3) were among the few recurrent genomic alterations, as well as TERT promoter mutations (63%, n=34/54 tested). Ex6::Ex17 variant samples were associated with greatest frequency of TP53 mutations (50% vs 21-31% for other fusion variants; p=0.101) and exclusively associated with PTEN mutations (19% vs 0%; p< 0.001). PDL1+ expression was associated with Ex6::Ex17 (25% vs 17% ‘other’, 2% Ex6::Ex16, and 0% Ex4::Ex2; p=0.002), yet Ex6::Ex17 TMEs had relatively low abundance of pro-inflammatory cell types, which were most abudance in Ex6::Ex16 tumors. SFTs from CNS-related sites had a trend toward a lower prevalence of TERT promotor mutations compared to non-CNS (36% vs 67-71%; p=0.12).

Conclusion: We molecularly characterized the largest known cohort of real-world SFT patient samples and identified new relationships between genomic alterations, tumor site, and NAB2::STAT6 fusion variant type. While confirming the association of thoracic SFTs with Ex4::Ex2 variants, we find that Ex6::Ex16 and Ex6::Ex17 variant subgroups have unique molecular profiles, including a notable difference in PDL1+ frequencies, which warrant further investigation of these fusion variant SFT subpopulations in order to develop new therapeutic approaches including the potential for immune checkpoint inhibitors.